Diabetic retinopathy is a common complication of diabetes. It affects millions people and is a leading cause of blindness in the industrialized countries. In diabetic retinopathy, there are two major pathological changes responsible for vision loss, diabetic macular edema (DME) and retinal neovascularization. DME can occur at early stages of diabetic retinopathy. The breakdown of the blood-retina barrier (BRB) and subsequent retinal vascular hyper-permeability or leakage of macromolecules from the blood vessels into the retina are believed to be the major causes of DME. It has been shown that the over-expression of vascular endothelial growth factor (VEGF) in diabetic retina plays a key role in the BRB breakdown and development of DME. Angiostatin, a proteolytic fragment of plasminogen, is a potent angiogenic inhibitor. It is present at high levels in normal vitreous and retina. Decreased vitreous angiostatin levels are associated with diabetic retinopathy. Recently, we have identified a new activity of angiostatin, i.e., blocking plasma protein leakage from blood vessels into the retina. A single intra-ocular injection of angiostatin can significantly reduce vascular leakage in the retina of the oxygen-induced retinopathy (OIR) and streptozotocin (STZ)-induced diabetes rat models. The effect on vascular leakage requires substantially lower doses than that required for its anti-angiogenic activity. Moreover, we have shown that angiostatin down-regulates VEGF expression in the retina, suggesting that angiostatin may function as a specific inhibitor of VEGF expression, which may be responsible for its effect on vascular leakage. The central hypothesis of this proposal is that angiostatin can have therapeutic effect in the treatment of DME. In this project, we propose the following studies to explore therapeutic potential of angiostatin in the treatment of DME. 1) To determine if trans-sclera deliveries of angiostatin can reduce retinal vascular leakage in diabetic rats. This study will develop and optimize a peri-ocular delivery route for angiostatin and determine its effect on retinal vascular leakage in diabetic rats. The optimized peri-ocular delivery route will be used for the sustained delivery of angiostatin. 2) To determine if sustained release of angiostatin can induce prolonged reduction of retinal vascular leakage in diabetic rats. Angiostatin will be formulated into polymer microspheres or pellets which will be implanted into the peri-ocular space to achieve sustained release. The long-term efficacy of the angiostatin-polymer on retinal vascular leakage will be evaluated. Our long-term goal is to develop a new, non-invasive and effective treatment for DME using natural peptides. This project will test the feasibility using angiostatin as a representative and will contribute to the development of a new drug therapy for DME.